sign in sign up
<<
Home , Vinorelbine

2617

Dose Escalation of Oral in Combination with Estramustine in Hormone-Refractory Adenocarcinoma of the Prostate

1 Niklas J. Mackler, M.D. BACKGROUND. The primary objective of the current study was to identify the 2 Rodney L. Dunn, M.S. tolerable dose level of oral vinorelbine when given in combination with estramus- 1 Beth Hellerstedt, M.D. tine to men with hormone-refractory prostate cancer (HRPC). The secondary 1,2 Kathleen A. Cooney M.D. objectives were to describe the toxicities of the combined regimen in patients with 1 Judith Fardig, P.A.-C. HRPC and to estimate the efficacy of oral vinorelbine in combination with estra- 1 Karin Olson, P.A.-C. mustine based on the prostate-specific antigen (PSA) response. 1,2 Kenneth J. Pienta, M.D. METHODS. Thirty-three patients with HRPC were treated on a 28-day cycle with 1,2 David C. Smith, M.D. estramustine at a dose of 140 mg orally 3 times a day on Days 1-3 and 8-10. Vinorelbine was given orally on Days 2 and 9. The initial dose of vinorelbine was 1 Division of Hematology/Oncology, Department of 50 mg/m2 and was escalated to 70 mg/m2 using the time-to-event continual Internal Medicine, University of Michigan School of reassessment method. Medicine, Ann Arbor, Michigan. RESULTS. Three of 17 patients experienced dose-limiting toxicity at the 70 mg/m2 2 University of Michigan Comprehensive Cancer dose level of oral vinorelbine. One patient experienced dose-limiting toxicity at a Center, University of Michigan, Ann Arbor, Michi- dose of 60 mg/m2 and no dose-limitig toxicities were reported at the 50 mg/m2 gan. dose. The overall response rate by Ն50% reduction in PSA was 17.2%, (95% confidence interval, 5.9-35.8%). CONCLUSIONS. Oral vinorelbine at doses of 70 mg/m2 may be safely combined with estramustine. The combination appears to have modest activity in men with advanced prostate cancer. The trial design employed the time-to-event continual reassessment method, which potentially allows for rapid accrual, a more complete assessment of toxicities, and a larger fraction of patients to be treated at an effective dose. More active regimens are needed to further evaluate the utility of this design in patients with prostate cancer. Cancer 2006;106:2617–23. © 2006 American Cancer Society.

KEYWORDS: prostate cancer, Phase I trials, biochemical resistant, clinical trials, adult medical oncology.

Supported in part by Grant 2P30 CA 46592-14 from the National Cancer Institute and the National denocarcinoma of the prostate is the second leading cause of 1 Institutes of Health and a grant from GlaxoSmith- Adeath from neoplasia in men. Roughly 50% of men will recur with Kline. advanced disease after definitive local therapy and most are effec- tively treated with androgen ablation. The duration of response to Presented in part at the 39th annual meeting of the androgen ablation is limited, however, with most patients developing American Society of Clinical Oncology, Chicago, Illinois, May 30–June 3, 2003. disease recurrence with hormone-refractory prostate cancer (HRPC). Responses to secondary hormonal therapies are typically short-lived Address for reprints: David C. Smith, M.D., 7302 and remain only a temporary option. In general, for CCGC 0946, 1500 E. Medical Center Dr., Ann HRPC has been considered to be palliative, although recent studies Arbor, MI 48109; Fax: (734) 615-2719; E-mail: have shown improvements in survival and quality of life.2–5 [email protected] A number of chemotherapy agents that block the Received November 14 2005; revision received apparatus have been evaluated in combination or as single agents in January 19 2006; accepted February 2 2006. HRPC. Estramustine (Emcyt; Pfizer, Groton, CT), a nor-nitrogen mus-

© 2006 American Cancer Society DOI 10.1002/cncr.21927 Published online 11 May 2006 in Wiley InterScience (www.interscience.wiley.com). 2618 CANCER June 15, 2006 / Volume 106 / Number 12 tard linked to an estrogen, has been used in combi- -based cytotoxic chemotherapy or radiation to nations with and to create Ն50% of the total bone marrow or if they had evidence complementary inhibition of microtubule assembly of brain metastasis or untreated spinal cord compres- and function.6–8 Estramustine binds to microtubule- sion. No prior malignancy except for in situ carci- associated proteins, inhibiting microtubule assembly. noma, nonmelanoma skin cancer, or adequately Vinorelbine (Navelbine; GlaxoSmithKline, Philadel- treated malignancy that had been inactive for Ͻ3 phia, PA), a , inhibits mitosis at meta- years was allowed. Patients with preexisting neuropa- phase through interaction with tubulin. In contrast to thy ՆGrade 2, active gastrointestinal disease, or a dis- other vinca alkaloids, vinorelbine appears to bind order that altered gastrointestinal motility or absorp- preferentially to the mitotic spindle, with lesser effects tion were also excluded. The Institutional Review on the in neural structures.9 Board of the University of Michigan Medical School In a Phase II trial of 25 patients with metastatic (Ann Arbor, MI) reviewed and approved the trial and prostate cancer using the combination of intravenous the informed consent document. Written informed vinorelbine and estramustine, 9 patients had a Ͼ65% consent was obtained from all patients before enroll- decline in prostate-specific antigen (PSA), whereas ment in the study. Data and safety monitoring were 10 patients had stable disease. Toxicities were mini- conducted under the criteria of the National Cancer mal and included alopecia, /vomiting, and Institute (NCI)-approved plan for the University of anorexia.10 A second Phase II trial of this combination Michigan Comprehensive Cancer Center. demonstrated a Ն50% decline in PSA in 15 of 21 patients with HRPC. The most frequent adverse events Trial Design were edema and thromboembolic complications.11 All patients were treated and followed as outpatients Oral vinorelbine is a newer formulation. A Phase II at the University of Michigan Comprehensive Cancer evaluation of oral vinorelbine in locally advanced or Center. The trial utilized a time-to-event continual metastatic demonstrated response rates reassessment method (TITE-CRM) design13–15 for the of 30%. The main adverse events were neutropenia dose escalation of vinorelbine. TITE-CRM was used to and gastrointestinal disorders. The efficacy and toxic- monitor the trial by estimating the rate of dose-limit- ity profiles of the oral formulation appeared to com- ing toxicity (DLT) while optimizing the number of pare favorably with those of intravenous vinorelbine.12 patients treated at effective doses and maintaining We conducted a Phase I trial to determine the open trial enrollment. The primary endpoint was to tolerable dose of oral vinorelbine in combination with identify the oral vinorelbine level associated with tox- oral estramustine on an intermittent schedule in pa- icity in 30% of patients when given in combination tients with HRPC. The secondary objective of the with oral estramustine. Secondary endpoints included study was an initial assessment of the efficacy of this determining the toxicities, response rate, time to dis- combination. ease progression, number of cycles of response, and time to treatment failure. MATERIALS AND METHODS Patients remained on primary androgen ablation Patient Eligibility during chemotherapy. Treatment consisted of 28-day Eligible patients had a histologic diagnosis of adeno- cycles with estramustine at a dose of 140 mg orally 3 carcinoma of the prostate with measurable (bidimen- times a day on Days 1-3 and 8-10 and vinorelbine sional) or evaluable (bone scan or rising PSA) disease. orally on Days 2 and 9. The estramustine dose was All patients underwent nonsteroidal antiandrogen held constant, whereas the vinorelbine dose level was withdrawal with a demonstrated rise in PSA after with- assigned using the TITE-CRM algorithm. The initial drawal. Those with PSA as their only evidence of pro- dose of vinorelbine was 50 mg/m2 based on the tox- gressive disease had to have a value Ն4 ng/mL. Eligi- icity estimates from prior studies. When a patient be- bility criteria included a granulocyte count Ն1500 came eligible for enrollment, the DLT for each dose cells/mm3, a platelet count Ն100,000 cells/mm3, bili- level was estimated based on our prior expectations of rubin Յ1.5 times the institutional upper limit of nor- toxicity and on the trial experience up to that time, mal, aspartate aminotransferase (AST) or alanine ami- and the highest vinorelbine dose with an estimated notransferase (ALT) Յ2.5 times the institutional upper DLT Յ30% was assigned to that patient. The experi- limit of normal, creatinine Յ1.2 times the institutional ence of prior patients, who had not completed the upper limit of normal, an Eastern Cooperative Oncol- 90-day acute toxicity observation window and who ogy Group (ECOG) performance status Յ2, age Ն18 had not experienced a DLT, was weighted proportion- years, and an estimated life expectancy Ն12 weeks. ally. The dose could not be escalated until at least 180 Patients were excluded if they had received prior vinca days of cumulative observation of acute toxicity in Oral Vinorelbine in Prostate Cancer/Mackler et al. 2619 patients enrolled at a particular dose level. Dose esca- surable and evaluable disease and/or a PSA Ͻ0.1 lation was restricted to 1 level between patients but ng/mL for at least 6 weeks. A partial response (PR) was there was no restriction on the number of levels that a Ն50% decrease from baseline in the sum of the the dose could be reduced. Doses could be decreased products of perpendicular dimensions of all measur- based on the accumulated toxicity data at any time able lesions and/or a Ն50% decrease in PSA from when the algorithm indicated that the 30% threshold baseline for at least 6 weeks. The response after each for toxicity rate had been exceeded. course was defined as a decrease of at least 50% from After completion of 3 cycles, patients who the maximum PSA value before that course of chemo- achieved a Ն50% reduction in the PSA from baseline therapy was started. Progression for the purposes of were observed until the PSA increased to Ն50% of the reinitiating therapy in patients who had responded nadir value, at which point they were eligible for re- was defined as a 50% increase in PSA from the lowest treatment. Patients who had Ͻ50% decrease in PSA value measured during or after the prior course of were continued on therapy until they had evidence of treatment. At all other times, progression was defined disease progression, toxicity, or patient/physician de- as a 25% increase in PSA from baseline, measured at 2 cision to discontinue therapy. Accrual for the study successive levels at least 4 weeks apart (in those with a continued until 30 evaluable patients were entered, nadir PSA Յ0.1 ng/mL, the level had to be Ն0.5 ng/mL with evaluable defined as patients completing at least by the second measure) or a 25% increase in the sum 1 cycle of chemotherapy or experiencing a DLT. of products of measurable lesions over the smallest sum observed, or appearance of any new lesion. For Dose Adjustments for Toxicity patients with bone-only disease, new lesions on bone No dose adjustments were made for an absolute neu- scan were required to meet the definition of progres- trophil count (ANC) of Ն1500 cells/m3 on the day of sion. Stable disease was defined as not meeting crite- treatment. If the ANC was between 1500 and 1000 ria for CR, PR, or progression. cells/m3 the vinorelbine dose was reduced by 50%. If the ANC was Ͻ1000 cells/m3, vinorelbine and estra- Statistical Analysis mustine were held. Patients were retreated after 1 At the end of the trial the posterior mean of the dose- week if the ANC reached the threshold values. Estra- toxicity parameter, ␣, and, correspondingly, the DLT mustine would then be given on Days 15-17 and vi- at each dose level, were calculated using the logistic norelbine given on Day 16. If retreatment criteria were dose-toxicity model and exponential prior distribution not met, estramustine and vinorelbine were held until on ␣ that were used to conduct the trial. The recom- the next cycle. If chemotherapy could not be admin- mended dose level for future study was based on istered for 3 consecutive weeks then treatment was choosing the highest dose level that met the target discontinued. Patients who had their dose reduced toxicity rate of 30%. All analyses were performed using could have the dose escalated back to full dose if SAS software (SAS Institute Inc., Cary, NC). subsequent ANC nadirs were Ն1500 cells/m3. Dose adjustments for vinorelbine were also made for he- RESULTS patic toxicity. For a bilirubin level of 2.1 to 3.0 mg/dL, Patient Characteristics 50% of the prior dose was given, and Ͼ3.0 mg/dL 25% Thirty-three eligible patients were registered from of prior dose was given. Dosing was also held for a February 10, 2002 to November 11, 2003. Patient char- Grade 2 or greater neurotoxicity. There was no dose acteristics are summarized in Table 1. All patients modification for estramustine. If a patient was unable were male and 97% were white. ECOG performance to tolerate estramustine they were to be withdrawn status was 0-1. The median age was 66 years (range, from the study. 49-82 years). Thirty-one had detectable cancer by ei- ther bone scan, CT, or both. All patients had hormone- Assessment of Toxicity and Response refractory disease with elevated PSA and a median DLT was defined using National Cancer Institute number of prior hormonal therapies of 2 (range, 1-4 Common Toxicity Criteria (version 2.0), as Grade 4 therapies). The majority (67%) had received prior che- hematologic toxicity lasting Ն7 days, any episode of motherapy (median number of regimens, 1; range, 0-3 febrile neutropenia, or ՆGrade 3 nonhematologic tox- regimens). Ten patients (30%) had previously received icity. At the completion of therapy, bone scan or com- and 12 (36%) had received estramustine. puted tomography (CT) scans were obtained if posi- tive on pretreatment evaluation and also at the time of Toxicity progression by PSA criteria. A complete response (CR) Of 33 patients enrolled in the study, 31 were evaluable was defined as a complete disappearance of all mea- for toxicity (Table 2). Patient 1 completed the first 2620 CANCER June 15, 2006 / Volume 106 / Number 12

TABLE 1 TABLE 3 Patient Characteristics Dose Level and Dose-Limiting Toxicity

Characteristic Value Prior DLT No. of DLT/ Posterior DLT Level Dose Estimate Patients Estimate Total no. of patients enrolled 33 Total no. of patients evaluable 31 1 30 10% 0/0 3.9% Median no. of cycles per patient 3 (range, 1–7) 2 40 15% 0/0 6.6% Median age in years 66 (range, 49–82) 3 50 20% 0/5 9.6% Median ECOG performance status 0 (range, 0–1) 4 60 25% 1/9 13% Sites of disease (no. of patients) 5 70 30% 3/17 16.8% Bone only 13 CT only 4 DLT: dose-limiting toxicity. Both bone scan and CT 14 PSA only 2 Median PSA 82.7 (range, 6.8–2558) Ͻ Prior local therapy (no. of patients) urinary obstruction and Grade 2 elevated func- 2 Radical prostatectomy 15 tion tests (at a dose of 70 mg/m ). His symptoms were Definitive RT 11 due to disease progression and were not believed to be Radical prostatectomy w/salvage RT 8 DLTs. One patient developed an episode of congestive Other 2 heart failure and fatigue (at a dose of 70 mg/m2) after Prior systemic therapy (no. of patients) Median no. of hormonal modalities 2 (range, 1–4) abruptly discontinuing his diuretic therapy. Two pa- Median no. of chemotherapy modalities 1 (range, 0–3) tients developed thrombosis, 1 mentioned above with No. of patients on vaccine trial 4 a myocardial infarction and another with a deep vein Palliative RT (no. of patients) thrombosis who continued therapy with anticoagula- External beam 9 tion until disease progression. Toxicities by dose level

ECOG: Eastern Cooperative Oncology Group; CT: computed tomography; PSA: prostate-specific anti- are summarized in Table 3. gen; RT: radiation therapy. Dose escalation and the associated toxicities at each level and for each patient are described in Table 4. Six patients were enrolled at the starting vinorelbine TABLE 2 2 Toxicity (Grade 3/4)* dose of 50 mg/m . The dose was then increased to 60 mg/m2 beginning with the seventh patient. The dose Lymphopenia 3 was escalated to 70 mg/m2 after Patient 11. After the Neutropenia 2 initial dose increase to 70 mg/m2, 3 patients devel- Neutropenia/fever 0 oped DLT. Therefore, when the dose for Patient 17 was 1 AST/ALK-P elevation 1 assigned, the TITE-CRM algorithm estimated that the 2 Nausea/vomiting 1 70 mg/m dose level had exceeded the target toxicity Fatigue 1 level and Patient 17 was assigned to a dose of 60 1 mg/m2. The next 4 subjects (Patients 17-20) were SOB 2 treated at the 60 mg/m2 dose, after which the vinorel- Hyperkalemia 1 Clotting 2 bine dose for the final 13 subjects was reescalated to 2 Incontinence 1 70 mg/m without further dose modification during the course of the trial. AST: aspartate aminotransferase; ALKP: alkaline phosphatase; SOB: shortness of breath. * Toxicity was assessed using National Cancer Institute Common Toxicity Criteria (version 2.0). Response Of the 33 patients enrolled, 29 were evaluable for a response (Table 5). The 2 patients who were not evalu- cycle but never returned for follow-up. Patient 12 re- able for toxicity were not evaluable for response. Pa- ceived 1 dose of vinorelbine and died unexpectedly 2 tient 8 developed dose-limiting atrial fibrillation and weeks later of a hemorrhagic cerebrovascular accident nausea during his first cycle of therapy. Patient 21 that was determined to be unrelated to therapy. DLTs developed rapidly progressive disease with ureteral included 1 patient with atrial fibrillation and nausea obstruction after 1 cycle of therapy and was subse- (dose level of 60 mg/m2), 2 patients with vomiting quently taken off study. His rapid disease progression (both at a dose of 70 mg/m2), and 1 patient with a required intervention and interruption of therapy. myocardial infarction (at a dose of 70 mg/m2) who was Neither patient had adequate therapy administered to found to have previously unsuspected, severe 3-vessel allow assessment of response. coronary artery disease. Another patient developed Five patients (17.2%; 95% confidence interval, 5.9- Oral Vinorelbine in Prostate Cancer/Mackler et al. 2621

TABLE 4 35.8%) had a PSA response during the course of their Dose Escalation and Toxicities therapy and entered the observation stage per proto- col. The median duration of observation before re- Dose Level, Non-DL Toxicity mg/m2 Patient No. (Grade 3-4)* DL Toxicity quiring the resumption of therapy was 48 days (range, 14-66 days). One patient experienced another PSA re- 50 1 (Not evaluable) sponse after being restarted on therapy and under- 2 Lymphopenia went a second period of observation, ultimately devel- 3 oping disease recurrence and progression on further 4 Lymphopenia 5 therapy. The second patient withdrew after entering 6 the observation phase as he had relocated to another 60 7 state. The third was restarted on therapy but withdrew 8 Atrial fibrillation after developing urinary obstruction despite develop- Nausea ing another PSA response. One failed to return for 9 10 further therapy due to his debilitated condition at the 11 time of disease progression, and the last patient re- 17 ceived an additional 4 cycles of therapy and then 18 withdrew from the study due to symptoms of fatigue. 19 The median number of cycles administered to 20 DVT 70 12 (Not evaluable) each patient was 3 (range, 1-7 cycles). At a median 13 Vomiting follow-up of 9 months, 13 patients (45%) had died. 14 Vomiting The median survival of all patients enrolled was 18.6 15 Myocardial infarction months. 16 21 22 Anemia, DISCUSSION diarrhea, SOB Therapy for HRPC remains palliative and the efficacy 23 of a regimen needs to be balanced against the toxicity. 24 Our aim in designing an oral regimen given in an 25 intermittent fashion was to minimize toxicity and to 26 27 Lymphopenia, emphasize the quality of life for patients in this set- 14,15 neutropenia ting. We elected to use the TITE-CRM for the dose 28 Pain escalation of oral vinorelbine. The TITE-CRM assumes 29 a simple model for the time to occurrence of toxic 30 response as a function of dose, and allows information 31 32 from all patients enrolled in the trial to be employed 33 SOB when assigning a new patient to a dose level. This method has several potential advantages over the “tra- DL: dose level; DVT: deep vein thrombosis; SOB: shortness of breath. ditional” 3 or 6 subjects per cohort design that is used * Toxicity was assessed using National Cancer Institute Common Toxicity Criteria (version 2.0). for Phase I studies. First, it allows subjects to be continuously re- TABLE 5 cruited throughout the trial, without recruitment Response pauses between dose levels. The duration of Ͼ 1.5 years to complete enrollment for this study reflects the Ն50% PSA Response 5 Response by vinorelbine dose restricted criteria inherent in our trial and the popu- 50 mg/m2 0 lation base available to participate at our institution. 60 mg/m2 2 Second, the target toxicity rate for the trial is variable. 70 mg/m2 3 The traditional Phase I design has a fixed implied rate Stable disease 7 of 33%. For this study, we chose 30% as the target rate. Progressive disease 17 Patient withdrawal from study 7 Third, the TITE-CRM design is potentially more accu- No. of patients entering observation 5 rate at determining the maximum tolerated dose Median duration of observation period 48 days (range, 14–66 days) (MTD) than the traditional design. An analysis with 500 simulations based on the posterior toxicity rates PSA: prostate-specific antigen. estimated for each of the 5 dose levels from our study (Table 3) shows that the traditional design correctly identifies the 70 mg/m2 dose as the MTD only 51.4% of 2622 CANCER June 15, 2006 / Volume 106 / Number 12 the time. In contrast, 500 simulations of the TITE- ously received estramustine and 30% had received CRM design using the same toxicity estimates found docetaxel. The other difference that could account for that the 70 mg/m2 dose level was correctly identified the decreased response rate is the route of adminis- as the MTD 100% of the time. The improved level of tration. However, the oral form of vinorelbine com- precision does not appear to be based on the in- pared favorably with the intravenous formulation in creased sample size for a TITE-CRM trial. Whereas the advanced breast cancer patients.12 Of the 29 patients traditional design would also accrue 30 patients if 6 evaluable for response, only 5 patients went on to the patients were needed for each of the 5 dose levels, the observation phase, with 3 resuming therapy when average sample size for the traditional design over the they developed evidence of progression by PSA crite- 500 simulations was 18.9 subjects with a mode of 18. ria. To test the impact of the increased sample size on the Other oral regimens that have been employed in precision of the MTD estimate, we repeated the 500 the treatment of HRPC have shown more promising TITE-CRM simulations using only 18 patients. The results. A Phase II trial combining oral cyclophospha- correct MTD was identified 97.2% of the time. Fourth, mide, diethylstilbestrol, and prednisone showed that a larger proportion of patients are able to be treated at 15 of 36 patients (42%) evidenced a decline in PSA of doses that are potentially therapeutic. This should Ն50%. Quality of life (QOL) evaluations from 17 pa- allow for better response estimates at the MTD dose. tients showed a significant improvement in respond- Conversely, there is the risk that the TITE-CRM-de- ers and no decline in QOL in nonresponders.18 Two signed trials may expose more patients to doses above Phase II trials of oral and estramustine have the MTD. Our findings suggested otherwise, because shown response rates ranging from 22% to 50% by PSA the regimen was well tolerated, with minimal toxicity. criteria.19,20 Therefore, the overall strategy of using However, timely reporting of toxicities is crucial to oral chemotherapy in this setting is reasonable, and avoid potential unnecessary risk. Finally, the TITE- further exploration of this approach is warranted. CRM design incorporates information for the entire time that a subject is enrolled in the study. An ap- Conclusions praisal of cumulative toxicities is not usually consid- The results of this Phase II study suggest that this ered with a traditional design, which assesses DLT combination has minimal activity in pretreated pa- only during the first cycle of therapy. Of the 4 DLTs tients with HRPC. The combination was well tolerated, from our study, 1 did not occur until the second cycle with little significant toxicity. Given the minimal level of therapy. of activity, this regimen is unlikely to provide signifi- This method performed well in this trial. The pri- cant benefit to men with prostate cancer despite the mary goal of the study was to determine a tolerable low overall rate of toxicity. Other combinations or dose of oral vinorelbine combined with estramustine. novel agents will be required to further test the utility We had estimated a toxicity rate of 30% at the 70 of this trial design. mg/m2 dose based on experience with these agents individually. The occurrence of DLT at this dose was REFERENCES lower than anticipated (Table 4). The combination of 1. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. oral vinorelbine and estramustine did not have any CA Cancer J Clin. 2004;54:8–29. significant increase in immediate or cumulative toxic- 2. Tannock IF, Osoba D, Stockler MR, et al. Chemotherapy ity. The recommended Phase II dose of oral vinorel- with plus prednisone or prednisone alone for bine is 70 mg/m2 in combination with estramustine. symptomatic hormone-resistant prostate cancer: a Cana- dian randomized trial with palliative end points. J Clin On- Seventeen patients in the trial were treated at the col. 1996;14:1756–1764. proposed Phase II dose with an overall response rate 3. Kantoff PW, Halabi S, Conaway M, et al. Hydrocortisone of 14%. The response rate is disappointing when com- with or without mitoxantrone in men with hormone-refrac- pared with trials utilizing intravenous vinorelbine in tory prostate cancer: results of the Cancer and Leukemia combination with estramustine in HRPC. Of 4 trials Group B 9182 study. J Clin Oncol. 1999;17:2506–2513. 4. Petrylak DP, Tangen CM, Hussain MH, et al. Docetaxel and employing this combination, the PSA response rates estramustine compared with mitoxantrone and prednisone 10,11,16,17 (Ն50% decline) ranged from 24% to 71%. Prior for advanced refractory prostate cancer. N Engl J Med. 2004; chemotherapy was an exclusion criterion in all the 351:1513–1520. trials of intravenous vinorelbine and this may account 5. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus pred- for the difference in response rates. Many of the 29 nisone or mitoxantrone plus prednisone for advanced pros- tate cancer. N Engl J Med. 2004;351:1502–1512. patients evaluable for response were heavily pre- 6. Speicher LA, Barone L, Tew KD. Combined antimicrotubule treated, having received a median of 2 hormone mo- activity of estramustine and taxol in human prostatic carci- dalities and 1 ; 36% had previ- noma cell lines. Cancer Res. 1992;52:4433–4440. Oral Vinorelbine in Prostate Cancer/Mackler et al. 2623

7. Hudes GR, Greenberg R, Krigel RL, et al. Phase II study of improvements in the continual reassessment method for estramustine and , two microtubule inhibitors, in phase I studies. Stat Med. 1995;14:1149–1161. hormone-refractory prostate cancer. J Clin Oncol. 1992;10: 14. O’Quigley J, Shen LZ. Continual reassessment method: a 1754–1761. likelihood approach. Biometrics. 1996;52:673–684. 8. Seidman AD, Scher HI, Petrylak D, Dershaw DD, Curley T. 15. Cheung YK, Chappell R. Sequential designs for phase I clinical Estramustine and vinblastine: use of prostate specific anti- trials with late-onset toxicities. Biometrics. 2000;56:1177–1182. gen as a clinical trial end point for hormone refractory 16. Smith MR, Kaufman D, Oh W, et al. Vinorelbine and estra- prostatic cancer. J Urol. 1992;147:931–934. mustine in androgen-independent metastatic prostate can- 9. Binet S, Chaineau E, Fellous A, et al. Immunofluorescence cer: a phase II study. Cancer. 2000;89:1824–1828. study of the action of navelbine, and vinblastine 17. Borrega P, Velasco A, Bolanos M, et al. Phase II trial of vinorel- on mitotic and axonal microtubules. Int J Cancer. 1990;46: bine and estramustine in the treatment of metastatic hor- 262–266. mone-resistant prostate cancer. Urol Oncol. 2004;22:32–35. 10. Carles J, Domenech M, Gelabert-Mas A, et al. Phase II 18. Hellerstedt B, Pienta KJ, Redman BG, et al. Phase II trial of study of estramustine and vinorelbine in hormone-refrac- oral , prednisone, and diethylstilbestrol tory prostate carcinoma patients. Acta Oncol. 1998;37: for androgen-independent prostate carcinoma. Cancer. 187–191. 2003;98:1603–1610. 11. Sweeney CJ, Monaco FJ, Jung SH, et al. A phase II Hoosier 19. Pienta KJ, Fisher EI, Eisenberger MA, et al. A phase II trial of Oncology Group study of vinorelbine and estramustine estramustine and etoposide in hormone refractory prostate phosphate in hormone-refractory prostate cancer. Ann On- cancer: a Southwest Oncology Group trial (SWOG 9407). col. 2002;13:435–440. Prostate. 2001;46:257–261. 12. Freyer G, Delozier T, Lichinister M, et al. Phase II study of 20. Sumiyoshi Y, Hashine K, Nakatsuzi H, Yamashita Y, oral vinorelbine in first-line advanced breast cancer chemo- Karashima T. Oral and oral etopo- therapy. J Clin Oncol. 2003;21:35–40. side for the treatment of hormone-refractory prostate can- 13. Goodman SN, Zahurak ML, Piantadosi S. Some practical cer. Int J Urol. 2000;7:243–247.